Epic: Prebuild transparency

[easyCZ]

Summary

Improve how we surface prebuild information, including status, duration and actionable errors to users
The main focus of this epic is on surfacing information about the "trigger" of a prebuild, and what consequence that trigger had.

Context

In the past we focussed on increasing prebuild reliability technically:

• Epic: Improve reliability of prebuilds and prebuild logs #7812
• Epic: Prebuilds stability #10361

But the perceived instability - and source of a big chunk of inbound user requests - is caused by uncertainty about if, how and why not a prebuild was triggered by some user action. This also affects Self-Hosted as the question "is my SCM correctly setup" is a frequently brought up during/after installation and can trivially answered from such a "prebuild trigger" view.

Also, this uncertainty will only increase when it comes to introducing more options for configuring prebuilds in the context of Usage Based Pricing.

Value

Clear and understandable prebuild statuses will increase trust in Prebuilds and as result in Gitpod as a product. It will enable users to optimize their prebuilds to be able to get the most out of Gitpod.

Acceptance Criteria

• Prebuild UI (and APIs) communicate prebuild information consistently
• Prebuild "triggers" are discoverable in the UI

Growth Area

• Acquisition
• Onboarding
• Exploration
• Expansion

Persona(s)

No response

Hypothesis

No response

In scope

No response

Out of scope

No response

Complexities

No response

Press release

No response

Issues

• triggers UI:

  • Show prebuild configuration state on the project prebuilds page  #7010
  • Record a "prebuild trigger" for every webhook event #10764
  • Display "prebuild trigger" in the UI #10765
  • Better instrument incoming webhooks via logging #7938

• observability:

  • [server] add a metric for received webhook events #9170
  • Create Grafana dashboard for Prebuilds metrics #10587

[jldec]

Added to roadmap.

[jldec]

Updated description with partially completed sub-issues and marking as done for Q3.